What Google announced
On 18 November 2026, Diego Rivas of Google DeepMind and Pavan Kumar of Google Labs published Genie 4 is now available. They say Google is rolling out a new Project Genie experience to eligible Google AI Ultra subscribers. The update unlocks first-time access to Genie 4, new upgrades to Genie 3, a Project Genie agent that helps users build their vision, and a community gallery to discover, explore, and remix worlds.
With Genie 4, which Google calls its most realistic and interactive world model yet, users can prompt mid-session to change a world they created, such as shifting the weather or adding a character. For the first time, Google says, Genie 4 pairs those visuals with real-time native audio that adapts as the scene unfolds, from ambient soundscapes to character dialogue. The official labs.google/projectgenie page repeats the Ultra eligibility and Pro waitlist, and describes Project Genie as an experimental research prototype for subscribers aged 18 and over.
Who can try it, and what else changed
Eligible Google AI Ultra subscribers can try the new experience starting the day of the announcement, and Google AI Pro members can join the waitlist on the same page. The post does not list a dollar price for Ultra, and it does not define which countries or accounts are eligible. This article therefore does not invent those details.
Genie 3 remains in Project Genie's model picker for longer runs or panorama-seeded spaces. Today's Genie 3 updates, as written in the post, include sessions of up to 3 minutes per generation (up from 60 seconds), the option to link multiple worlds, 360-degree panorama image grounding, and real-time in-world events from natural-language prompts. The 3-minute figure is stated for Genie 3, not for Genie 4.
The agent, the gallery, and stated limits
The new Project Genie agent takes a natural-language description and, Google says, orchestrates multimodal models to create initial visuals and custom background scores. Users can also bring Gemini models into a session for suggestion chips or in-world events. A community gallery offers curated and community worlds, with the option to remix and publish. All video and audio generated in Project Genie includes SynthID watermarking, and inputs and outputs go through safety screenings. Google still describes Genie 3 and Genie 4 as world models in frontier research, with work ongoing on camera latency, physical interactions, real-time audio precision, and long-term scene permanence.
Go to the source
Sources & further reading
Based on the official documentation below. Features and access can change; check the provider before making a decision.
Was this page useful?
Help us make our coverage more useful. This is feedback on the page, not a rating of the tool.
Go a little further